Commands, agents and skills I've made or collected for use with Claude Code
Command
There are BugBot, Cursor or Devin comments on this PR. I want you to fix and address these comments from these bots, and then after you have pushed the fixes, I want you to reply to those comments and then mark them resolved.
Command
Read the meeting transcript in $ARGUMENTS and clean it up so that it uses proper grammar and eliminates filler words like "um", "uh", etc. When you are done it should be easier to read, but without changing the actual content being communicated in any way beyond minor grammatical corrections and a bit of sentence…
Command
Use $ARGUMENTS and thoroughly review the code in this repository, looking for correctness, security, performance, adherence to community best practices, use of best-in-class supporting packages and libraries, reducing code duplication, opporutinities to simplify or streamline, and any areas where an artful use of…
Command
Commit all work as a series of atomic, logically sequenced commits. Each commit should represent one coherent change that can be understood, reviewed, and reverted independently.
Command
Diagnose and prune bloated Claude Code context. Supports treat, reload, guard mode, and doctor.
Command
Deep multi-language code review with specialist sub-agents.
Command
Parse receipts and generate a filled expense report spreadsheet.
Command
Diagnose and resolve Alertmanager alerts using NixOS tools.
Command
CI tests are failing for this PR. Use cpp-pro to diagnose and resolve this issue, then push your fixes to the PR and monitor the CI test results using gh until you observe that everything passes correctly. If any further problems should be observed, repeat this diagnose, resolve, push and monitor process until…
Command
Please analyze and fix the GitHub issue: $ARGUMENTS.
Command
I am trying to install the Home Assistant integration $ARGUMENTS but it is showing me this output.
Command
Read $ARGUMENTS and rewrite this file into a more properly formatted English document. Don't change any of the meaning, just give it better structure and form: paragraphs, grammatical corrections, punctuation, capitalization where needed, etc.
Command
Think deeply to analyze the following query, construct a well thought out plan of action based on the following context, and then carefully execute that plan step by step.
Command
There are still more flaky tests, as shown at $ARGUMENTS.
Command
Use the forge skill to perform a comprehensive multi-model collaborative workflow on the following problem.
Command
Act like gravity for my idea. Your job is to pull it back to reality. Attack the weakest points in my reasoning, challenge my assumptions, and expose what I might be missing. Be tough, specific, and do not sugarcoat your feedback.
Command
Remember the following.
Command
Please analyze this codebase and create a CLAUDE.md file, which will be given to future instances of Claude Code to operate in this repository.
Command
Install and configure a service with nginx, monitoring, and secrets.
Command
Write in John Wiegley's authentic voice. Use when drafting technical articles, blog posts, essays, or any written material that should read as if John wrote it himself. Captures patterns from 1,100+ posts spanning 1992-2026 across both technical (newartisans) and personal (johnwiegley) blogs. Focuses on technical…
Command
Add a lefthook.yml file for running $ARGUMENTS code formatting, building with no warnings, running tests, linting, code coverage checking, etc., on pre-commit.
Command
/heavy Ensure the logit comparison test passes for $ARGUMENTS from /pos/weights.
Command
Write all of these to a GitHub-flavored Markdown document, using the suggestion feature of GitHub, so that I can easily copy and paste these into GitHub review comments wherever appropriate.
Command
Remember the following.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: